
\begin{table}[h]
\caption{Heterogeneous Treatment Effects - Rebel Presence (Trust Outcomes) with Block Bootstrapped Standard Errors}
\begin{center}
\begin{threeparttable}
\scalebox{.8}{
\begin{tabular}{l@{} D{.}{.}{3.5}@{} D{.}{.}{3.5}@{} D{.}{.}{3.5}@{} D{.}{.}{3.5}@{} D{.}{.}{3.5}@{}}
\toprule
 & \multicolumn{1}{c}{Trust} & \multicolumn{1}{c}{Trust} & \multicolumn{1}{c}{Trust} & \multicolumn{1}{c}{Trust} & \multicolumn{1}{c}{Perceived} \\
 & \multicolumn{1}{c}{Index} & \multicolumn{1}{c}{(Safe)} & \multicolumn{1}{c}{(Intent)} & \multicolumn{1}{c}{(Info)} & \multicolumn{1}{c}{NPA Sympathy} \\
\cmidrule(lr){2-2} \cmidrule(lr){3-3} \cmidrule(lr){4-4} \cmidrule(lr){5-5} \cmidrule(lr){6-6}
 & \multicolumn{1}{c}{(1)} & \multicolumn{1}{c}{(2)} & \multicolumn{1}{c}{(3)} & \multicolumn{1}{c}{(4)} & \multicolumn{1}{c}{(5)} \\
\midrule
Treated           & 0.17 & 4.86 & 6.24^{***} & 2.26 & -6.86^{**} \\
 					& (0.10) & (3.01) & (2.37) & (1.37) & (2.71) \\
NPA Presence 2015 & 0.23^{**} & 2.95 & 8.72^{***} & 4.32^{**} & -5.11^{**} \\
 					& (0.10) & (2.90) & (2.77) & (1.94) & (2.50) \\
Treated*NPA       & -0.57^{***} & -17.10^{***} & -19.14^{***} & -5.04^{***} & 12.50 \\
 					& (0.13) & (5.59) & (3.95) & (1.94) & (7.65) \\
Base Outcome      & 0.40^{***}  & 0.33^{***}   & 0.35^{***}   & 0.28^{***} & 0.41^{***} \\
                  & (0.04)      & (0.06)       & (0.07)       & (0.03)     & (0.03)     \\
\midrule
R$^2$             & 0.17 & 0.12 & 0.13 & 0.11 & 0.16 \\
Observations      & \rlap{612} & \rlap{635 } & \rlap{628 } & \rlap{628}        & \rlap{636}        \\
Clusters          & \rlap{18}          & \rlap{18}           & \rlap{18}           & \rlap{18}         & \rlap{18}         \\
\bottomrule
\end{tabular}}
\end{center}
\begin{tablenotes}[flushleft]
\scriptsize{\item \sym{*} \(p<.10\), \sym{**} \(p<.05\), \sym{***} \(p<.01\)} 
\item The results in this table correspond to those shown in Table \ref{tab:HetEffects_NPAsub}. Models in this table use the panel specification denoted in equation 1 and include a dummy variable for missing baseline. All models also include rank fixed effects for the blocking variables, though they do not include station fixed effects because NPA Presence is measured at the municipality (station) level. Standard errors are clustered and block-bootstrapped at the station-level.}
\end{tablenotes}
\end{threeparttable}
\label{tab:HetEffects_NPAsub_bootstrapped}
\end{table}
